Cordless light switches

Yeah, I've had these upstairs at work for a few years and had no bother, I would highly recommend, not cheap but makes the job a lot easier.

There is nothing (electrically) dodgy about combining two lighting circuits onto one MCB - assuming the total load does not exceed the capacity of the MCB.

There is a user issue, that a fault on the one now larger circuit will trip all lights rather than just one those on one story.

Sometimes it might be possible to rewire a part of the upstairs circuit to remove the landing light from that circuit entirely, and connect it to the downstairs one instead.

In message <t3r9uj$d70$ snipped-for-privacy@dont-email.me, at 10:59:48 on Thu, 21 Apr

Even if both are spurs? That was the issue my electrician had.

The main safety risk here is the stairs, and whether they are connected to upstairs, downstairs (or both at the same time) one RCD tripping will plunge them into darkness anyway.

It's gaining the necessary access which is a problem. And I definitely don't want to start cutting new channels in the walls. I suppose one could run a cable up an outside wall, but I usually try to remove unsightly outside cables, not install more of them. And that would be a third spur, dear liza.

Lighting circuits are not rings, so they are essentially just lots of spurs in a tree. I suppose you could connect a single feed going from fitting to fitting but it's certainly not required and there's absolutely nothing wrong with branching as much as you like as long as the overall load is OK. You're very unlikely to hit loading limits nowadays with mostly LEDs.

In message snipped-for-privacy@esprimo.zbmc.eu>, at 13:08:21 on Thu, 21 Apr

The issue the electrician had was having the "root" of two of separate such spurs(/trees) connected into one breaker in the consumer unit.

Maybe he's wrong about the regs, but he did check with a much more senior colleague.

Meanwhile I've looked at the upstairs landing light switch and it has three cables.

One is "going down", and is almost certain to be the other end of the two cores "going up" from the downstairs hall switch.

There's another pair of cores "going up".

And the third is 'straight through', and is probably the spur from the consumer unit to the upstairs tree.

I don't think allows us to reinstate a conventional two-way switch (because you'd need three cores between the two floors) but it could mean that we don't need to rummage around much in the attic to find a live pair to connect a smart-controller to.

There's no disagreement about that.

I don't really understand the use of the term "spur" when applied to a lighting circuit, which is traditionally wired as radial. Hence you can have as many branches as you like at any point in the circuit.

Yup, and even spread over two circuits there is no guarantee that the bit you need to navigate safely will be adequately lit - especially if the stairs change direction at any point.

(Mine do, hence I have non maintained emergency lighting at both top and bottom!)

With the upstairs circuit, much of thee wiring is often accessible from the loft. So it will then come down to how many cables run between ground and first floors.

Even with a socket circuit you can legitimately run a spur from the circuit origin - i.e. three wires connecting to the MCB and other bus bars. One for each end of a ring, and the third for the spur.

With a lighting circuit, the same applies you can branch wherever you like, including at the circuit origin.

I wonder if there is some communication at crossed purposes happening here?

You mean there is a cable not connected to the switch?

Yes that could be the power feed from the CU to the upstairs circuit - probably running straight to a ceiling rose.

If the circuits were merged, then you can do two way with fewer wires:

That's how they work.

They should be fine, as they are only available as single switches. No chance of two getting out-of-sync.

Not every 'socket circuit' is a ring of course.

NICEIC had a comment about that when they did one of my inspections. I had put the smokes on with the upstairs lights to the same MCB (two cables at the MCB). I was told it was wrong but when asked for a reg they just said "we do not like it"

No riskier than a power cut. The number of times it has been mentioned that stairs and lights not working/failing are a risk to life I have seen on this newsgroup is unbelievable.

The main cause of stair accidents is alcohol and not a lack of lighting.

Ikea do something like that - hubless switches that can control their own wireless bulbs. I've got a couple on the hall/stairs/landing.

The switch contains the battery - last about 2 years.

I've a few Xiaomi LED battery lights with motion sensors dotted about - takes AA rechargable batteries that need charging every 6 months or so.

The older MEM MCBs even had a wiring diagram on them telling you on which order to place the three cables. The cables were terminated behind a plate and not is a screw clap like modern MCB

In message <t3s6m3$q12$ snipped-for-privacy@dont-email.me, at 19:10:11 on Thu, 21 Apr

Indeed so, that was part of his explanation. However he said two spurs despite being fewer wires wasn't OK.

It's possible they both had "the wrong answer", but they knew what the question was.

At the moment none are connected (he capped them off), but yes, not

The closest ceiling rose is directly above, and the one whose arrangements are under discussion.

Yes, that was what I was expecting the outcome to be (if we can in fact merge the two lighting circuits). But noting the compatibility issues with RCDs.

Meanwhile, if it's still OK to use the earth as a third conductor (as in their "standard two way circuit"), then we have three conductors in each direction.

In message <t3sdfe$hqh$ snipped-for-privacy@dont-email.me, at 21:06:07 on Thu, 21 Apr

That's probably where my installation is at the moment.

The first house I had with RCDs, the vast majority of the 'power cuts' were false alarms by the RCD.

It's particularly applicable when the RCD trips in the middle of the night, the householder is upstairs and the consumer unit is downstairs.

We need to cross each bridge as we come to it. I'm not proposing DIY brewing as an aspect of this particular thread!

In message <t3s4v2$bqn$ snipped-for-privacy@dont-email.me, at 18:40:50 on Thu, 21 Apr

It means one cable being the root of the tree at the consumer unit. Not a ring.

In my house the upstairs light illuminates the whole stairway, but the downstairs light only that half.

The loft is difficult to access at all, and full of very hazardous-looking insulation and general detritus. If can find someone to give me a grant, I'd like (someone - it's beyond my DIY ambitions) to clear it all out and re-insulate to modern standards.

I've been looking into that.

I had that - likely due to too many PCs and general gadgets, all with their own (leaky) filters on the power supplies, pushing it near the limits, so any tiny disturbance was enough to push it over.

Scrapping the RCD and replacing all the MCBs that it fed with RCBOs shared the leakage out and we have had no false trips since.

Even that is ok, so long as any overload condition is prevented by the nature of the loads on the spurs.

Normally the circuit protective device (Fuse/MCB/RCBO) is responsible for providing both overload current protection, and fault current protection.

While fault current protection must always be provided at the origin of the circuit, there are cases where overload protection can be delegated elsewhere.

The classic example of this is a normal spur from a 32A ring circuit. The maximum current carrying capacity of a single length of 2.5mm^2 T&E is less than the trip threshold of the MCB. However the rules for an unfused spur limit the load to one socket accessory (single or double). That means there is no way to cause an overload on the spur cable. A fault current however (say a short at the end of the spur) will cause the MCB to open, and adequately protect that single spur cable. (you can prove this if you calculate the maximum prospective fault current and then put that into the adiabatic equation)

The rules don't change with the location of the spur - so a ring with two spurs is "ok" (not ideal, and messy in wiring terms, and often physically difficult to wire), but technically ok. So by extension, two spurs at origin is ok, and then removing the ring and just leaving the two spurs is also ok. Now you have in effect made a radial circuit with a branch at the origin, and delegated the overload protection to the load end of the circuit.

So for example you have a couple of 13A immersion heaters, you could wire each in 2.5mm T&E, and connect both cables to the same 32A MCB. The

Or possibly, as Adam commented, the various professional bodies / trade associations gold plate their own "rules" on top as "best practice", and hence forbid practices that BS7671 itself deems ok, or is silent about.

If both circuits are merged, they are then served from the same RCD, and that problem goes away.

Does the circuit as a whole have earthing? (pre 1966, lighting circuits were not required to be earthed).

If it does not[1], then the question is simplified to "is it ever acceptable to re-purpose a protective conductor as something else?" The answer to that is a qualified "yes" so long as the conductor is part of a multi wire cable, and not a loose "single" and that is is correctly over marked at both ends.

That is not a spur - just a branch in a radial.

The only time you would have a genuine spur on a radial, is where you run a backbone connection in a larger CSA cable, but then connect loads to it using a smaller cable. Say a 32A radial wired in 4mm^2 T&E, and then individual socket drops in 2.5mm^2 T&E. The MCB would then only provide fault protection for the socket drops, not overload protection.

Fair enough... some lofts become "no go" areas with increasing time, and decreasing enthusiasm!

But then you have the issue that one RCD tripping will lose all lights in the house